R (35 terms)
Rademacher Complexity
Measures a model’s ability to fit random noise; used to bound generalization error.
Intermediate
RAG
Architecture that retrieves relevant documents (e.g., from a vector DB) and conditions generation on them to reduce h...
Intermediate
Random Variable
Variable whose values depend on chance.
Advanced
Rank
Number of linearly independent rows or columns.
Advanced
ReAct Pattern
Interleaving reasoning and tool use.
Advanced
Reality Gap
Differences between simulated and real physics.
Advanced
Recall
Of true positives, the fraction correctly identified; sensitive to false negatives.
Intermediate
Recurrent Neural Network
Networks with recurrent connections for sequences; largely supplanted by Transformers for many tasks.
Intermediate
Red Teaming
Stress-testing models for failures, vulnerabilities, policy violations, and harmful behaviors before release.
Intermediate
Reflection Prompting
Asking model to review and improve output.
Intro
Reflex Agent
Simple agent responding directly to inputs.
Advanced
Regularization
Techniques that discourage overly complex solutions to improve generalization (reduce overfitting).
Intermediate
Reinforcement Learning
A learning paradigm where an agent interacts with an environment and learns to choose actions to maximize cumulative ...
Intermediate
ReLU
Activation max(0, x); improves gradient flow and training speed in deep nets.
Intermediate
Representation Learning
Automatically learning useful internal features (latent variables) that capture salient structure for downstream tasks.
Intermediate
Reproducibility
Ability to replicate results given same code/data; harder in distributed training and nondeterministic ops.
Intermediate
Residual Connection
Allows gradients to bypass layers, enabling very deep networks.
Intermediate
Responsible AI
A discipline ensuring AI systems are fair, safe, transparent, privacy-preserving, and accountable throughout lifecycle.
Intermediate
Restricted Boltzmann Machine
Simplified Boltzmann Machine with bipartite structure.
Intermediate
Retrieval Prompt
Prompt augmented with retrieved documents.
Intro
Reward Hacking
Maximizing reward without fulfilling real goal.
Advanced
Reward Model
Model trained to predict human preferences (or utility) for candidate outputs; used in RLHF-style pipelines.
Intermediate
Reward Shaping
Modifying reward to accelerate learning.
Advanced
Rigid Body Dynamics
Motion of solid objects under forces.
Advanced
Risk Model
Quantifying financial risk.
Intermediate
Risk Register
Central log of AI-related risks.
Intermediate
Risk Stratification
Grouping patients by predicted outcomes.
Intermediate
RLHF
Reinforcement learning from human feedback: uses preference data to train a reward model and optimize the policy.
Intermediate
Robotics
Field combining mechanics, control, perception, and AI to build autonomous machines.
Advanced
Robust Alignment
Maintaining alignment under new conditions.
Advanced
Robust Control
Control that remains stable under model uncertainty.
Intermediate
ROC Curve
Plots true positive rate vs false positive rate across thresholds; summarizes separability.
Intermediate
Role Prompting
Assigning a role or identity to the model.
Intro
Rotary Positional Embeddings
Encodes positional information via rotation in embedding space.
Intermediate
RRT
Sampling-based motion planner.
Advanced